Echo Identification of Vessel Activities Using HF Radar and Relevant Threat Evaluation

• Main Authors: Yu-JenChung, 鍾育仁
• Other Authors: Laurence L. H. Chuang
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/54078032465040639724

博士 === 國立成功大學 === 海洋科技與事務研究所 === 104 === There is a growing need to deal with threats to national security coming from the ocean. Compared to conventional radar, high-frequency (HF) ground wave radar transmits low-power radar waves that can extend the detection distance up to several hundred kilometers with little attenuation. Moreover, HF radar has various advantages, such as active, continuous monitoring and near real-time performance. It is thus worth developing a vessel detection algorithm based on the HF radar system for large area surveillance. The present study aims to achieve three objectives. The first is to analyze the local environmental noise to understand its distribution and effects. The second is to extract ship echoes from the HF radar system’s sea-returns by applying a simple procedure to the cross-spectra series of the signals of the SeaSonde HF radar system. An adaptive detection technique (ADT) is used to build a threshold surface which can be adaptive to the local environmental noise. In this work the ship information retrieved from the HF radar data was validated against the automatic identification system (AIS) data. The final aim is to explore the potential use of this system for maritime management, with a focus on threat evaluation of non-cooperative targets. Since environmental noise may contain the complex irregular variations, this study first explores this noise to understand the regional characteristics and system limits. Appropriate observations in range and time intervals are derived from the spatial and temporal variations of the environmental noise, which can help develop a better detection scheme. The results show that the background noise levels during the nighttime are significantly higher than those during the daytime. Similar to the regularities of the ionosphere, the diurnal variations of time series can be derived by spectral analysis. Furthermore, the results of the environmental noise analysis show that the better time period and scope for vessel detection is from 02:00 to 07:00 (UTC) and the range cells from 4 to 12, within which the targets can be easily identified. Based on ADT, a two-dimensional moving average filter is applied to build an adaptive threshold surface in the region of interest to extract ship motion data. Actual information for the vessels collected from the AIS database closely matched that derived from the HF radar sea echoes, which confirms the vessel signal detection capabilities of this HF radar. In addition, the main advantage of ADT is the decision process applied to find the optimal window, which may change based on the properties pof the environmental noise. The process of optimal window determination in ADT is not a tradeoff check, and this can be decided according to the spatial and temporal characteristics of the environmental noise. Furthermore, the unidentified targets, which may be the non-cooperative ones, are determined by ADT and checked against the AIS. The proposed threat evaluation procedure is a preliminary idea which not only provides an initial and rapid classification of possible targets, but also helps the command system to take the appropriate actions using the related information. The threat levels can thus be computed to obtain the related information, which can be regarded as part of the decision support system to better understand the reaction time of the command system with regard to maritime security.